Intelligent control and coordination of an Au-tonomous Mobile Robot (AMR) fleet is a challenging problem. Optimization of the fleet performance using centralized control is feasible only for small fleets. Decentralized control, on the other hand, is feasible also for large fleets and features flexibility and robustness. We emphasize this flexibility and robustness and consider intelligent control as the capability of a system of mobile robots to manage and handle as many maintenance tasks as possible at any time. We focus on task assignments and use entropy in a dualistic way. For task assignments, maximization of potential entropy is sought for the set of tasks supported by available AMRs. By contrast, task performance is considered by minimization of the expected actual entropy of the AMR's task execution. Distributed fleet control is proposed with a contract negotiation protocol utilizing the entropy maximization and minimization measures.

Maintenance tasks represent a potential area for applying multi-purpose Autonomous Mobile Robots (AMRs). Intelligent control and coordination of such a system is challenging and optimization methods are feasible only for small fleets. Decentralized control can provide flexibility and robustness, which are better applicable also for large fleets, though with less guaranteed performance. Our focus is on flexibility and robustness in task scheduling and task assignments and we use entropy as an indirect performance criterion for coordination, both at the system level (maximize entropy) and at the AMR level (minimize entropy). As a distributed coordination scheme, we use a modified contract negotiation protocol. We show preliminarily the feasibility of our approach with simulation results.

This paper presents a concept for next generation collaborative assembly stations for manufacturing industry. It applies a collaborative mobile platform, AI and 5G technologies. Feasibility of the 5G and Edge processing was studied in this paper. Remote operations and data processing were realized with 5G New Radio wireless connection together with edge processing in order to support low latency communication for data transfer.

Industry 4.0 is forming the manufacturing industry towards more agile and networked operations enriched by exploitation of new technology. Manufacturing is also one of the key verticals defined by the 5G Infrastructure Public Private Partnership (5G PPP) aimed at digital transformation in industrial communication. In this paper, we will present the current status of the VTT’s (VTT Technical Research Centre of Finland Ltd) Industry X Smart Manufacturing Testbed, which enables Finnish industrial companies to test and experiment new production methods and technologies in an industrial-like environment. The testbed utilizes technologies varying from robotics to 5G and beyond connectivity, and from international data spaces (IDS) to extended forms of reality (AR, VR, MR) and digital twins. The testbed is built in a hierarchical manner. Inter-connected factory level testbeds comprise of factory process level processes that are grouped into use case families. These processes include several proofs of concept demonstrators, which utilize different technologies available in the testbed. In this paper, three recent proof-of-cencepts (PoC) are briefly presented to illustrate the next-generation manufacturing possibilities using the latest communication technologies available.

In this paper, we present a method for planning the sensing features when the geometrical representation of the target object is known. The presented method is a synthesis form and can be used in several measurement applications in robotics. Sensing planning is an important issue when the measurement data is sparse, includes a lot of noise or there are tight time-requirements. The criteria for selecting the measurement locations and orientations is a posteriori error covariance matrix of the parameters to be estimated. The presented approach is verified by simulation tests in the case of work object location. The proposed method has been implemented into a commercial ConiferRob-robot software

Calibration of coordinate systems between cameras and robots, i.e. hand-eye-calibration, is of key importance in sensor based robotics. We present a method which is based on using a very simple, conic calibration object, and results in fast and simple calculations. It originates from the pose estimation algorithms and contributes to earlier methods robustness for the location of the calibration object and simplicity from the point of view of image processing. Our method is also based on the Bayesian model, thus it also provides estimates for the spatial uncertainties, in the form of covariances of the pose parameters. The algorithm has been tested by simulations and practical experiments.

A holonic multi-robot cell is a cell specifying a system based on autonomous and cooperative units called holons. This paper presents a method using Bayesian estimation to model uncertainties in a multi-robot cell. Using sensor fusion with several robots and combining all the information together gives new possibilities for planning the sensing operations, e.g. to minimize the size of locating measurements. If the robots and work object are accurately modelled (noise model and surface forms), an iterative planner can be used to plan a minimal task sequence for the treatment of the work object. Further analysis of the remaining uncertainties is based on observing the direction of the eigenvectors and the respective eigenvalues of the error covariance matrix.

The traditional view of monitoring is extended from the execution-time validity checking of actions towards dynamic plan-validity checking and forecasting of risky situations. Monitoring for intelligent robots is recognized as an independent task, deeply embedded in the dynamics of planning and execution. The solution presented is based on very general assumptions of plans and planning, and is thus widely applicable within control systems over the whole bandwidth of reactivity. JANUS, an experimental monitoring shell based on the concepts of situation grids and world state aspect objects as generic monitors, has been implemented.

A general description of a control system for a prototype of an intelligent and skilled paper roll manipulator (PRM) is given. The elementary operations related to both manipulations and environment observations are described, and integrating these with a hierarchically organized planning-executing-monitoring architecture (PEM) is discussed. The kernel of PEM architecture is a goal-oriented module consisting of these three principal parts. The motion control of the paper roll manipulator is divided into free motion (gross motion) and force-constrained motion (fine motion) phases. Gross motion control takes the paper roll from the starting position to the vicinity of the goal position, whereas the fine motion control is used in placing the paper roll against other paper rolls and the support. For coarse recognition of the environment a 3D laser device is used. The accurate position of paper rolls in the picking phases is measured with the aid of an ultrasonic sensor system, and the final force constrained placement is carried out with force sensing based on measurements of hydraulic pressures of the manipulator system.

This paper describes a knowledge-based planning and execution system concept for sensor-guided manipulation. The main purpose of the approach is to offer a flexible and easy prototyping method to build a sensor-guided manipulation system and to integrate multi-sensors. Qualitative knowledge about manipulation is represented in a rule-base system and actual control commands are greater in the functional modules of the manipulator and the sensors by the use of object models and sensor models. The proposed idea was applied to paper roll manipulation. The simulation system implemented was composed of a rule-based action planner, simulators for paper roll manipulator, force and vision sensors and a notice-board. Simulation of the paper roll stacking task showed that the concept is effective and applicable.

Implementing a control system for an autonomous machine is a challenging task. Several techniques have to be applied, such as task planning, hierarchical and/or distributed control, and advanced sensing techniques. In addition, to be useful these various techniques have to be integrated into a system that has to operate more or less in real-time. The authors present a control scheme based on hierarchically organized planning-executing-monitoring-cycles which is used to solve some of the problems related to real-time control of an autonomous machine. The implementation is also presented in which the control system is applied in a pilot system based on an industrial robot.>